Automatic BIM component extraction from point clouds of existing buildings for sustainability applications

• A method of automatic building geometry extraction from point cloud is proposed.
• Components of building envelope were recognized and modeled from point clouds.
• The proposed method was validated by testing on three as-is building data.

Building information models (BIMs) are increasingly being applied throughout a building's lifecycle for various applications, such as progressive construction monitoring and defect detection, building renovation, energy simulation, and building system analysis in the Architectural, Engineering, Construction, and Facility Management (AEC/FM) domains. In conventional approaches, as-is BIM is primarily manually created from point clouds, which is labor-intensive, costly, and time consuming. This paper proposes a method for automatically extracting building geometries from unorganized point clouds. The collected raw data undergo data downsizing, boundary detection, and building component categorization, resulting in the building components being recognized as individual objects and their visualization as polygons. The results of tests conducted on three collected as-is building data to validate the technical feasibility and evaluate the performance of the proposed method indicate that it can simplify and accelerate the as-is building model from the point cloud creation process.